哈里森图像数据集

HARRISON 是一个对 Facebook、Twitter、Instagram 社交媒体中的图片进行标签标注（tag）的图像数据集，包括 57383 张图片，每张图片平均 4.5 个标签标注，标签标注来自出现频率最高的 1000 个单词。 主题标签定义为附加在前缀字符“＃”上的任何单词，该单词在在线社交网络服务（SNS）（例如Facebook，Twitter和Instagram）中使用。随着在线社交网络的增长，标签通常用于总结用户帖子的内容并吸引关注者的注意。在社交媒体时代，推荐适当的标签是一项非常有趣和有用的任务。 我们介绍了用于图像主题标签推荐的新颖基准，称为HARRISON，或针对社交网络中的真实世界图像的HAshtag建议。HARRISON数据集是一个现实的数据集，由来自Instagram的57,383张照片和每张照片的平均4.5个相关主题标签（最少1个，最多10个）组成。每个图像的地面真相标签由1,000个最常用的标签组成，并根据频率排名结果以数字进行编码。 数据引文：

HARRISON is an image dataset for hashtag annotation of images from social media platforms including Facebook, Twitter, and Instagram. It contains 57,383 images, with an average of 4.5 hashtags per image, where the hashtags are selected from the top 1000 most frequently used terms.

A hashtag is defined as any word prefixed with the '#' character, which is widely used in online social network services (SNS) such as Facebook, Twitter, and Instagram. With the growth of online social networks, hashtags are commonly used to summarize the content of user posts and attract the attention of followers.

In the era of social media, recommending appropriate hashtags is an intriguing and valuable task. We present a novel benchmark for image hashtag recommendation named HARRISON, short for HAshtag Recommendation for real-world images in Social networks.

The HARRISON dataset is a real-world dataset composed of 57,383 photos sourced from Instagram, paired with an average of 4.5 associated hashtags per photo (ranging from a minimum of 1 to a maximum of 10). The ground-truth hashtags for each image are selected from the 1000 most frequently used hashtags, and encoded numerically based on their frequency ranking results.

Data Citation: